Abstract We have identified two novel tumor-specific subtypes of pancreatic cancer which we call ?classical? and ?basal- like?. Several findings illustrate the power of our findings. Our subtypes were not only prognostic in our own dataset but also independently validated the presence and prognostic significance of our subtypes in the recently published International Cancer Genome Consortium (ICGC) PDAC microarray dataset as well as the unpublished publically available data from The Cancer Genome Atlas (TCGA). Importantly, genes from the ?basal-like? factor, including laminins and keratins, were also consistent with basal subtypes previously defined in bladder and breast cancers, leading us to hypothesize that like both these cancers basal subtypes although more aggressive, are conversely more sensitive to cytotoxic therapies. We have also found that both genetically engineered mouse models and cell lines inadequately represent our classical subtype with only patient-derived xenograft (PDX) tumors showing representation of both. Finally we have also confirmed and strengthened our previous association between kinome activation profiles since our initial submission. Our two subtypes suggesting that novel kinase inhibitor approaches may be tailored to each subtype. Thus we will use novel proteomic approaches to assess the baseline activation state of the kinome in each of our subtypes and use PDX models to determine whether our subtypes are associated with response to novel kinases inhibitor and cytotoxic therapies. Furthermore to advance the translation of our findings we propose to begin to evaluate the association of our subtypes with treatment response in existing deidentified patient samples from two different institutions (Dana-Farber Cancer Institute and Indiana University).